Introduction
Myringoplasty remains a cornerstone procedure for re-establishing tympanic membrane integrity in patients with chronic middle-ear pathology. Its primary goals are to prevent recurrent otorrhoea, allow undisturbed water contact and to restore more efficient sound transmission. The earliest attempt to close a perforation was described by Berthold in 1878,Reference Berthold1 and Wullstein later refined this concept by introducing the term Type I tympanoplasty for reconstructions performed without ossicular chain manipulation.Reference Wullstein2 Since then, various surgical approaches—transcanal, endaural and retroauricular—and different grafting techniques such as underlay and overlay have been adopted, employing materials including fascia, perichondrium and cartilage. More recently, endoscopic techniques have increasingly challenged traditional microscopic approaches. Despite this technical diversity, the essence of the procedure remains unchanged: achieving a stable, functional tympanic membrane capable of protecting the middle ear and transmitting sound to the ossicles.
Most published series report high graft success rates accompanied by an improvement in the air–bone gap (ABG), although the extent of hearing gain differs considerably across patient populations.Reference Merchant and Rosowski3–Reference Sheehy and Anderson5 This variability has stimulated ongoing interest in identifying anatomical or clinical factors that influence functional outcomes. More recently, a study by Lerut et al.Reference Lerut, Pfammatter, Moons and Linder6 demonstrated pre-operatively a linear association between perforation size and conductive hearing loss across 0.5–4 kHz and described an “inverted V-shaped’’ ABG pattern characterised by greater deficits at low and high frequencies. Their findings also corroborated Rosowski’s observation that anterior versus posterior location has minimal impact on hearing, whereas perforations involving the umbo are associated with a more pronounced ABG.
The present investigation was conducted in a clinical setting with a long-standing culture of systematic documentation in otologic surgery. All procedures were performed according to the standards of Prof. Fisch, as outlined in the second edition of Tympanoplasty.Reference Fisch, May and Linder7 Within this framework, we evaluated 465 successfully healed myringoplasties, focusing specifically on post-operative audiometric changes. Our objectives were threefold: to characterise the overall post-operative hearing improvement and determine whether myringoplasties of posterior perforations yield audiometric outcomes that differ significantly from perforations located elsewhere on the tympanic membrane; and to address pitfalls in data acquisition and follow-up.
Materials and methods
Recruitment of patients
Using a prospective database (Innoforce ENT statistics, http://www.innoforce.com), a search was performed for all patients, who underwent a primary myringoplasty between 2001 and 2024. A total of 522 primary myringoplasty procedures were initially identified. All patients presented with a central tympanic membrane perforation, the majority had a history of previous ear infections with otorrhoea but their perforation was mostly dry at time of surgery. Additional inclusion criteria required at least two months of post-operative clinical follow-up, intact ossicular chain mobility and an intact tympanic membrane at the final follow-up visit. From the initial cohort, patients with insufficient audiometric data and incomplete follow-up, or reperforations and patients with severe mixed hearing loss lacking reliable bone-conduction thresholds were excluded. After applying these criteria, 442 patients who underwent 465 procedures with successfully healed myringoplasties were included in the final analysis.
Patient groups and follow-up
Patients were grouped as “posterior” when the perforation was located in the posterior part of the tympanic membrane and as “other than posterior” when the perforation was located in any other part of the tympanic membrane including subtotal perforations. Additionally, the patients were further divided into two groups based on the timing of their surgery: those who underwent the procedure before 2013, and those who had the procedure after 2013. A minimum follow-up of two months was established for all patients.
Analysis of the data
Demographic details, surgical records, operation diagrams, audiological measurements and follow-up information were collected from our database. For assessing the air conduction and bone conduction thresholds, the four-tone average of 0.5, 1, 2 and 4 kHz frequencies was used. Because 3 kHz thresholds are not routinely recorded in our service, they were estimated by extrapolating from the 2 and 4 kHz data when the difference between these two frequencies was less than 15 dB. The post-operative audiogram with the longest follow-up period was selected for the analysis. Data distribution was assessed and found to be non-normal; therefore, nonparametric tests were applied.
Statistical analysis was performed using the ENTStatistics software. The Wilcoxon signed-rank test was used for paired pre- and post-operative comparisons, while the Mann–Whitney U test was applied for comparisons between independent subgroups. Comparisons were made using mean ABG values for 0.5–4 kHz (0.5, 1, 2, 3, 4 kHz) frequency ranges. A p-value of less than 0.05 was considered statistically significant. This study was approved by swissethics (national Swiss ethics committee) with the BASEC Nr. 2025-02521.
Surgical technique
All operations were carried out by senior surgeons as well as junior otologists. Myringoplasties were performed in accordance with the principles outlined by Fisch, May and Linder in Tympanoplasty, Mastoidectomy and Stapes Surgery.Reference Fisch, May and Linder7 For small posterior perforations, an endaural approach with limited canalplasty was preferred; in endoscopic cases, the same approach was likewise selected for appropriately sized small perforations, whereas a retroauricular incision with canalplasty was used in all other types of perforations. Graft materials consisted of temporalis fascia in the retroauricular approach, tragal cartilage and pericondrium in the endoscopic approach and perichondrium in the endaural approach. In the endaural and retroauricular approach graft positioning was in an underlay technique.
In posterior perforations, the “swinging-door” modification was applied in most patients to allow circumferential freshening of the perforation margins (Figure 1A). In subtotal and anterior perforations, an anterosuperior canalplasty was achieved allowing the pullthrough of the graft without disturbing the anterior tympanomeatal angle and therefore avoiding blunting (Figure 1B).Reference Fisch and Linder8
Surgical techniques used for myringoplasty. (A) Swinging-door technique used for circumferential freshening of the perforation margins in posterior tympanic membrane perforations. (B) Elevation of the tympanomeatal flap with anterosuperior canalplasty to achieve secure graft fixation in subtotal and anterior perforations.
Use of artificial intelligence tools
During manuscript preparation, digital tools were employed to improve linguistic clarity and presentation quality of the text and figures. ChatGPT (OpenAI) and Gemini (Google) were used exclusively for language editing and figure enhancement. These tools were not involved in data acquisition, data analysis, scientific interpretation or the formulation of conclusions. The authors assume full responsibility for the scientific content, analyses and interpretations presented in this study.
Results
Within the 442 patients who underwent 465 myringoplasties, there was a near-equal sex distribution, with 229 men and 213 women with a mean age of 35 years (range, 4–84 years). The mean pre-operative ABG for the frequencies of 0.5 to 4 kHz was 19.1 dB (AG1), and the median value was 18.75 dB (Table 2). The maximum was at 4 kHz and the minimum at 2 kHz (Table 1). The average post-operative follow-up was 3.5 years (42 months; range, 2–264 months or 22 years; SD, 46 months). Only 9 per cent of all patients did have a short follow-up. Post-operatively, 53 per cent of patients demonstrated an ABG below 10 dB and 96 per cent achieved an ABG below 20 dB, with only one patient exceeding 30 dB. Overall, the mean ABG post-operatively measured 9.8 dB, and the median was 10 dB. Comparing individual frequencies the mean post-operative ABG measured 15 dB at 4 kHz and 4.5 dB at 2 kHz; 76 per cent of the cohort had an ABG under 10 dB at 2 kHz, compared with 25 per cent at 4 kHz. All data were evaluated separately for posterior perforations and for all other perforation types, and the cohort was additionally divided into two subgroups based on whether surgery was performed before or after 2013. Among the 70 patients with posterior perforations, the post-operative ABG across 0.5–4 kHz averaged 8.7 dB, whereas in all other perforation types the corresponding value was 10.2 dB (p = 0.0226). When stratified by surgical period, the mean post-operative ABG was 9.6 dB in procedures performed before 2013 and 10.8 dB in those carried out thereafter (p = 0.0349). The Glasgow plot (Fig. 2) demonstrates that most patients reached normal or near-normal air-conduction thresholds, with the best outcomes observed in patients whose contralateral ear was healthy.
Post-operative air–bone gap (AG2) outcomes across pure-tone averages and individual frequencies
Note: Values are expressed as percentages.
ABG = air–bone gap.
Pre-operative (AG1) and post-operative (AG2) air–bone gap values across frequencies (n = 465)
Note: Q1 and Q3 represent the 25th and 75th percentiles of the data, respectively. The interquartile range (IQR) is a measure of variability that describes the spread of the central 50 per cent of the distribution.
AG1 = pre-operative air–bone gap; AG2 = post-operative air–bone gap; SD = standard deviation.
Glasgow Benefit Plot (0.5–4 kHz). Post-operative air-conduction thresholds are shown as purple circles and pre-operative thresholds as orange stars. Hearing thresholds of the non-operated contralateral ear are shown for comparison. The green shaded area represents the range of normal hearing thresholds.
Figure 2 Long description
Title: Glasgow Benefit Plot (0.5-4 kHz). The horizontal axis label is Non-op ear AC [dB HL]. The horizontal axis shows values from 0 to 120. The vertical axis label is Op ear AC [dB HL]. The vertical axis shows values from 0 to 120. A legend shows Audiogram 1 and Audiogram 2. The plot contains many scattered points, with a dense cluster in the lower-left area where both axes are below about 40. Additional points extend across the plot up to around 100 on both axes. A diagonal reference line runs from near (0, 0) toward (120, 120). A second diagonal line runs parallel above it. A vertical reference line is drawn at about 40 on the horizontal axis. A horizontal reference line is drawn at about 30 on the vertical axis, with the text Normal hearing near the right side.
Analysing patients with a post-operative ABG greater than 20 dB, the majority presented with myringosclerosis within the remaining eardrum prior to surgery, a type B tympanogram post-operatively indicating the possibility of middle ear effusion or a stiff membrane and frequently a diseased contralateral ear.
Discussion
In a recent survey among German-speaking countries, a marked heterogeneity regarding diagnostic pathways for ear surgeries was observed among otology centres.Reference Morgenstern, Neudert, Zahnert and Lailach9 Only half of the centres performed speech audiometry routinely before tympanomastoid surgeries. For “simple” myringoplasties, a pre- and post-operative speech audiogram is rather indicated when a mixed hearing loss with documented inner-ear dysfunction is present. A routine speech audiogram, therefore, is not performed in most cases of myringoplasties. It would also add costs for healthcare overall. In cases of tympanic membrane perforation, pre-operative tympanometry typically yields a type B (flat) curve. However, the volume measurement may still be informative, as discussed by Epprecht et al.Reference Epprecht, Qingsong, Stenz, Hashimi and Linder10 A higher volume of more than 2 cc suggests good mastoid pneumatisation and thus a healthy mastoid, potentially obviating the need for additional computed tomography (CT) imaging in tympanomastoid procedures without cholesteatoma.Reference Epprecht, Qingsong, Stenz, Hashimi and Linder10 The influence of perforation size and location on the pre-operative ABG was examined by Lerut et al., who demonstrated a correlation between perforation size and hearing loss, as well as an additional negative effect when the umbo was involved. The perforation’s location, however, showed no significant impact.Reference Lerut, Pfammatter, Moons and Linder6 Many different surgical techniques have been described to close eardrum perforations, and different materials are widely used. More recently, the endoscopic approach challenges endaural or retroauricular skin incisions using the microscopic approach. Using the endoscope, many surgeons prefer to detach the tympanic membrane from the malleus handle to reach anterior and anterosuperior perforations. The graft is still placed as an underlay technique. Lotto et al. presented data supporting the malleus detachment techniqueReference Lotto, Fink, Stricker, Fernandez and Presutti11 obtaining a post-operative mean ABG of 11.1 dB. However, their follow-up time was only three months. Microscopic surgeons are often trained not to detach the remaining drum from the malleus to preserve sound transmission from the short process to the umbo. At our centre, tympanoplasties are performed using the Fisch technique in a uniform manner. Posterior perforations are addressed by an endaural approach (Figure 1A) or using the endoscope and a simple underlay graft using perichondrium or fascia. Cartilage is very rarely used. Kidney-shaped larger or subtotal perforations are predominantly closed using the microscopic retroauricular approach and using temporalis fascia (Figure 1B). Our first review in myringoplasties involving 106 patients operated by two senior surgeons with a mean follow-up of 26 months documented a post-operative ABG of 8.2 dB.Reference Pfammatter, Novoa and Linder12 Our current study involved four times the number of patients. Within the first 10 years, three senior surgeons operated between 50 and 70 patients each, whereas three junior surgeons in training contributed to a small number of the final outcome. Analysing the next 10 years, the number of patients per senior surgeons dropped to 23–64 cases each, whereas 10 junior otologists and fellows contributed to 2–20 cases each. The mean post-operative ABG with the contribution of the most senior surgeons reached 9.6 dB, whereas the last 10 years including more younger colleagues was slightly (p = 0.0349) worse averaging 10.8 dB. The difference was small indicating that involving younger colleagues to perform otologic surgeries should not lead to a worse outcome.
Most otology centres agree that a follow-up period of at least one year is necessary to adequately assess hearing outcomes, recurrence rates and complications in tympanomastoid surgery for non-cholesteatoma patients. However, patients with tympanic membrane perforations often return to their referring physician or feel sufficiently satisfied after the initial improvement and, therefore, do not attend the recommended scheduled visits. As a result, many reports on myringoplasty outcomes lack long-term follow-up data. Asking students or residents in training, most of them expect a post-operative ABG close to 0 dB and do not really distinguish between frequencies. Our 465 myringoplasties with a mean follow-up time of 3.5 years reached a post-operative ABG of 10.04 dB (0.5–4 kHz). Hearing thresholds below 10 dB were achieved in 53 per cent of patients, and thresholds below 20 dB were achieved in 96 per cent. Even with a successful myringoplasty, it is often not possible to achieve a fully healthy and functionally restored middle ear in all cases. The overall results seem even better, when only reporting from 0.5 to 3 kHz, as suggested by the American Academy of Otolaryngology–Head and Neck Surgery guidelines for the evaluation of treatment results for conductive hearing loss.13 In our series, the mean post-operative ABG would be calculated as 8.48 dB. Consequently, it becomes evident that closing the ABG at 4 kHz continues to pose a significant challenge. The published literature indicates considerable heterogeneity in outcome reporting. Functional improvements have been described as gain in ABG without referring to the final ABGReference Karela, Berry, Watkins and Phillipps14, Reference Gamra, Nacef, Abid, Harilga and Mbarek15 or improvements of air-conduction thresholds, some only report to 2 kHz, others do not mention the follow-up time frame.
This variability in reporting limits direct comparison between studies and may partly explain the wide range of reported functional outcomes following myringoplasty. In this context, the present study provides a detailed, frequency-specific evaluation of post-operative ABG distribution, allowing a more comprehensive assessment of hearing outcomes. In a large series of 1476 myringoplasties with a mean follow-up of 10 months, Berglund et al. reported a mean post-operative ABG of 11.8 dB.Reference Berglund, Olaison, Bonnard, Fransson, Hultcrantz and Florentzson16 In a series of 553 patients, Al Busaidi and Al Rahbi calculated a mean post-operative ABG of 12.9 dB with a 12 months follow-up but included only 0.5–3 kHz.Reference Al Busaidi and Al Rahbi17
In an interesting Finnish study that initially included 338 ears between 1986 and 2012, hearing outcomes across 0.5–4 kHz were reported for 302 surgeries at one-year follow-up and for 175 cases (51 per cent) at three years. The majority of procedures (84 per cent) were performed by a senior surgeon. The mean ABG was 9.2 dB at one year and 9.0 dB at three years post-operatively. After exclusion of cases with reperforation, the mean ABG at one-year follow-up improved to 8.1 dB. When stratified by perforation size, the mean post-operative ABG was 12.2 dB for large perforations and 8.6 dB for small perforations, with both values including patients with reperforation.Reference Iso-Mustajärvi, Dietz and Löppönen18 In our series, which included a comparable number of patients and a similar follow-up duration, ABG closure within 10 dB was achieved in 53 per cent of cases, closely matching the 56 per cent reported in the Finnish series. However, ABG closure within 20 dB was observed in 96 per cent of our patients, compared with 80 per cent in the Finnish cohort.
Pitfalls and enthusiasm in reporting
With the introduction of newer techniques—whether using artificial materials to close perforations or adopting methods that minimise incisions and visible scarring—initial reports often present outcomes that appear superior to those of earlier approaches. Such results, however, can be influenced by methodological factors, including shorter follow-up periods, omission of data at 4 kHz, exclusion of paediatric patients or a focus primarily on improvements in the ABG without comparing pre- and post-operative air- and bone-conduction thresholds. Studies focusing on intermediate- or long-term follow-up often struggle with high patient dropout rates, audiograms provided by referring ENT colleagues that contain inadequate masking, missing bone-conduction thresholds or poorly recorded curves. Newer techniques must withstand the test of time. Therefore, long-term follow-up—typically with a median of three to five years—is essential to validate the initial enthusiasm surrounding them. Careful reading of the materials and methods section, often overlooked as tedious and boring, becomes highly relevant for otologists who maintain and analyse their own databases. Young colleagues beginning their otology careers should start documenting their surgeries as early as possible and consistently evaluate their own outcomes, comparing them with the existing literature and engaging with more experienced colleagues. The best and most extensive database is only as valuable as the surgeons’ commitment to entering data consistently and prospectively.
What is already known on the subject:
• Myringoplasty is a standard procedure for closing tympanic membrane perforations and improving hearing.
• While graft success rates are high, functional audiological outcomes vary significantly in the literature, often due to inconsistent reporting standards and follow-up durations.
What this paper adds:
• This large-scale study of 465 primary myringoplasties shows that even after successful anatomical closure, a residual air–bone gap (ABG) of 8–12 dB should be expected, particularly at 4 kHz.
• It highlights that posterior perforations yield significantly better hearing results and emphasises the importance of long-term, frequency-specific reporting to avoid over-optimistic success rates.
Conclusions
Even in successful myringoplasty cases with an intact tympanic membrane, a residual ABG of approximately 8–12 dB across 0.5–4 kHz can be expected. Posterior perforations yield optimal results. ABG closure remains most challenging at 4 kHz. Overall, experienced surgeons tend to achieve slightly better hearing outcomes than junior surgeons in training. Therefore, junior otologists should continuously and prospectively document their surgical cases to monitor outcomes and refine their techniques. A common limitation across all studies is the high dropout rate, which may reach up to 50 per cent at three-year follow-up.
Author contributions
All authors have contributed significantly to this work. R.O., S.T. and Y.R. were involved in data acquisition and analysis; G.P. and C.S. contributed to scientific interpretation; TL supervised the study and performed surgical procedures.
Competing interests and source of funding
The authors declare no conflicts of interest and received no external funding for this study.
